| Message ID | 20190730172137.23114-14-eric.auger@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | VIRTIO-IOMMU device | expand |
On Tue, Jul 30, 2019 at 07:21:35PM +0200, Eric Auger wrote: > When translating an address we need to check if it belongs to > a reserved virtual address range. If it does, there are 2 cases: > > - it belongs to a RESERVED region: the guest should neither use > this address in a MAP not instruct the end-point to DMA on > them. We report an error > > - It belongs to an MSI region: we bypass the translation. > > Signed-off-by: Eric Auger <eric.auger@redhat.com> Reviewed-by: Peter Xu <peterx@redhat.com>
On Tue, Jul 30, 2019 at 07:21:35PM +0200, Eric Auger wrote: > When translating an address we need to check if it belongs to > a reserved virtual address range. If it does, there are 2 cases: > > - it belongs to a RESERVED region: the guest should neither use > this address in a MAP not instruct the end-point to DMA on > them. We report an error > > - It belongs to an MSI region: we bypass the translation. > > Signed-off-by: Eric Auger <eric.auger@redhat.com> something wrong with the subject here. > --- > > v9 -> v10: > - in case of MSI region, we immediatly return > --- > hw/virtio/virtio-iommu.c | 16 ++++++++++++++++ > 1 file changed, 16 insertions(+) > > diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c > index 8e54a17227..20d92b7ab0 100644 > --- a/hw/virtio/virtio-iommu.c > +++ b/hw/virtio/virtio-iommu.c > @@ -711,6 +711,7 @@ static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr, > viommu_interval interval; > bool bypass_allowed; > bool read_fault, write_fault; > + struct virtio_iommu_probe_resv_mem *reg; > > interval.low = addr; > interval.high = addr + 1; > @@ -743,6 +744,21 @@ static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr, > goto unlock; > } > > + reg = g_tree_lookup(ep->reserved_regions, (gpointer)(&interval)); > + if (reg) { > + switch (reg->subtype) { > + case VIRTIO_IOMMU_RESV_MEM_T_MSI: > + entry.perm = flag; > + return entry; > + case VIRTIO_IOMMU_RESV_MEM_T_RESERVED: > + default: > + virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_MAPPING, > + 0, sid, addr); > + break; > + } > + goto unlock; > + } > + > if (!ep->domain) { > if (!bypass_allowed) { > qemu_log_mask(LOG_GUEST_ERROR, > -- > 2.20.1
diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c index 8e54a17227..20d92b7ab0 100644 --- a/hw/virtio/virtio-iommu.c +++ b/hw/virtio/virtio-iommu.c @@ -711,6 +711,7 @@ static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr, viommu_interval interval; bool bypass_allowed; bool read_fault, write_fault; + struct virtio_iommu_probe_resv_mem *reg; interval.low = addr; interval.high = addr + 1; @@ -743,6 +744,21 @@ static IOMMUTLBEntry virtio_iommu_translate(IOMMUMemoryRegion *mr, hwaddr addr, goto unlock; } + reg = g_tree_lookup(ep->reserved_regions, (gpointer)(&interval)); + if (reg) { + switch (reg->subtype) { + case VIRTIO_IOMMU_RESV_MEM_T_MSI: + entry.perm = flag; + return entry; + case VIRTIO_IOMMU_RESV_MEM_T_RESERVED: + default: + virtio_iommu_report_fault(s, VIRTIO_IOMMU_FAULT_R_MAPPING, + 0, sid, addr); + break; + } + goto unlock; + } + if (!ep->domain) { if (!bypass_allowed) { qemu_log_mask(LOG_GUEST_ERROR,
When translating an address we need to check if it belongs to a reserved virtual address range. If it does, there are 2 cases: - it belongs to a RESERVED region: the guest should neither use this address in a MAP not instruct the end-point to DMA on them. We report an error - It belongs to an MSI region: we bypass the translation. Signed-off-by: Eric Auger <eric.auger@redhat.com> --- v9 -> v10: - in case of MSI region, we immediatly return --- hw/virtio/virtio-iommu.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+)